Skip to content
TheChatArchiveThe Chat Archive
Perplexitysonar-pro· PerplexityHugging Face LMArena Search Arena 24K

how much does search-GPT-4o API cost

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Perplexitysonar-reasoning-pro-high’s answer.
  1. User

    how much does search-GPT-4o API cost

  2. Perplexity Assistant

    GPT-4o API Pricing Overview

    OpenAI’s GPT-4o API pricing is based on the number of tokens processed—both for input (your prompts) and output (the model’s responses). Pricing is set per 1 million (1M) tokens, and there are different rates for the standard GPT-4o model and the lighter, cheaper GPT-4o-mini variant.

    Standard GPT-4o API Pricing

    Token TypePrice per 1M Tokens
    Input$2.50
    Output$10.00
    • Example: Generating a 900-word article (about 1,200 output tokens) would cost roughly $0.012 for the output alone on GPT-4o[3][7].

    GPT-4o-mini API Pricing

    Token TypePrice per 1M Tokens
    Input$0.15
    Output$0.60
    • GPT-4o-mini is designed for high-volume, cost-sensitive applications and is about 15 times cheaper than the standard GPT-4o model[3][7].

    Web Search Tool Pricing (Search-GPT-4o)

    If you use the web search capability (sometimes called "search-GPT-4o" or "gpt-4o-search-preview"), pricing is per 1,000 (1k) search calls and varies by context size:

    Context SizePrice per 1k Calls (GPT-4o)Price per 1k Calls (GPT-4o-mini)
    Low$30.00$25.00
    Medium$35.00 (default)$27.50
    High$50.00$30.00
    • The "search context size" refers to how much information is retrieved and processed per search. Higher context sizes cost more[2].

    Summary Table

    ModelInput (per 1M tokens)Output (per 1M tokens)Web Search (per 1k calls, medium)
    GPT-4o$2.50$10.00$35.00
    GPT-4o-mini$0.15$0.60$27.50
    GPT-4o-search-preview$2.50$10.00$35.00
    GPT-4o-mini-search$0.15$0.60$27.50

    Key Points

    • Both input and output tokens are billed—factor both into your cost estimates[3][7].
    • Web search calls are billed separately from token usage, based on the number of searches and the context size[2].
    • GPT-4o-mini offers significant cost savings for high-volume or less demanding use cases[3][7].
    • Pricing may vary slightly depending on the platform (OpenAI direct, Azure, or third-party resellers), but the above rates are standard as of April 2025[2][7][8].

    Example Calculation

    Suppose you make 1,000 API calls, each using 500 input tokens and generating 1,000 output tokens:

    • Total input tokens: 500 × 1,000 = 500,000 tokens
    • Total output tokens: 1,000 × 1,000 = 1,000,000 tokens

    Cost for GPT-4o:

    • Input: 0.5 × 2.50=2.50 = 1.25
    • Output: 1 × 10.00=10.00 = 10.00
    • Total: $11.25 (excluding any web search tool usage)

    Cost for GPT-4o-mini:

    • Input: 0.5 × 0.15=0.15 = 0.075
    • Output: 1 × 0.60=0.60 = 0.60
    • Total: $0.675

    Conclusion

    • GPT-4o API: 2.50per1Minputtokens,2.50 per 1M input tokens, 10.00 per 1M output tokens.
    • GPT-4o-mini API: 0.15per1Minputtokens,0.15 per 1M input tokens, 0.60 per 1M output tokens.
    • Web search tool (search-GPT-4o): $35.00 per 1,000 calls (medium context size).

    These rates make GPT-4o one of the most cost-effective advanced models for both general and search-augmented AI applications[2][3][7].

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.